Closed mhamjediers closed 1 year ago
Having the kmatch dependency and two-step process, I guess we just document the approach we'd take if we want standardization and do not implement that option.
I think we should even do this with the kmatch-dependency
Implemented as NORMalize
option (normalize is the better term I think), for standalone only. Always normalized with respect to the depvar
mean of bref
group and therefore changes with treat
levels and swap
(and, of course, bref
).
Closing.
Is implemented in the first setup of the main command; but shall we keep the variable? should be included also in the upper matching table?